Monday, 28 September 2026

Where technology leaders come to think out loud

Silvia Lehnis
Image: UBDS Digital

The VETTDD 50 · UK IT channel · 2026Rising leaders

No.49

Silvia Lehnis

Chief AI Officer, UBDS Digital · ubdsdigital.com ↗ · LinkedIn ↗

UBDS Digital’s chief AI officer argues that agentic AI projects succeed or fail on skills and operating models, and her practice now has a measured production result to show for it

group turnover, year to November 2025, up 11.8%
£30.95m
time saved on catalog reviews at Longmead Capital, per UBDS
20%
chief AI officer by
Jun 2026

Silvia Lehnis is chief AI officer at UBDS Digital, a London-based IT services firm that sells cloud, cybersecurity, data and AI work mainly to the public sector and regulated industries. She held the title by June 2026, having previously been the firm’s consulting director for data and AI, the role in which IT Pro quoted her in November 2025.

The firm behind the practice

UBDS Group Holdings, the parent, grew turnover 11.8% to £30.95m in the year to 30 November 2025, according to accounts filed at Companies House in August 2026, while integrating two 2024 acquisitions and setting up a delivery arm in India. In June 2026 LDC, the private equity arm of Lloyds Banking Group, took a minority stake. The accounts name AI governance and practical automation as growth areas, and UBDS holds ISO 42001 certification for AI management systems.

From pilot to production

The practice’s published work from the past 12 months is specific. In May 2026 UBDS described a reusable Databricks data platform for an unnamed public-sector client, which it says four teams have adopted. In September it published a case study on Longmead Capital, an investment firm, where it set up a five-stage method for adopting Anthropic’s Claude and moved portfolio catalog management into production. Longmead estimates this saves about 20% of the time spent on accuracy reviews and 60–70 minutes of editorial review per major catalog entry, according to UBDS.

An argument about skills

Lehnis’s public case is that agentic AI projects succeed or fail on the operating model around the technology. In June 2026, writing for UBDS and in a techUK guest blog, she set out three skills teams need: framing the right problem, managing organizational knowledge and managing agents. “When you can build at machine speed, the expensive mistake is now building the wrong thing quickly,” she wrote. She also argued that a person signing off every agent action does not scale.

In July 2026 she was one of 20 people shortlisted for CRN UK Woman of the Year. The winners are announced on 8 October.

Why it matters to the channel

For services partners, the harder sale is not an AI pilot but a client process running in production, with measured savings and the governance trail that public-sector buyers expect. That is the ground Lehnis works on. With LDC’s investment behind UBDS, she is building its AI practice around production results and a clear argument about the skills that make agents work.

Silvia Lehnis’s leader profile →The full 2026 list →